JBUON 2020; 25(1): 554-565ISSN: 1107-0625, online ISSN: 2241-6293 • www.jbuon.comEmail: editorial_office@jbuon.com

ORIGINAL ARTICLE

Corresponding author: Eylem Taskin, PhD. Department of Physiology, Faculty of Medicine, Nigde Omer Halisdemir University, Nigde, Turkey,Tel: +90 3882252594, Fax: +90 3882252582, Email:eylemtaskin@yahoo.comReceived: 14/06/2019; Accepted: 27/07/2019

Silencing HMGB1 expression inhibits adriamycin’s heart toxicity via TLR4 dependent manner through MAPK signal transductionEylem Taskin1*, Celal Guven2*, Salih Tunc Kaya3, Melda Sariman4, Zeliha Emrence4, Sema Sirma Ekmekci4, Neslihan Abaci4, Yusuf Sevgiler5, Handan Akcakaya6

1Department of Physiology, Faculty of Medicine, Nigde Omer Halisdemir University, Nigde, Turkey; 2Department of Biophysics, Faculty of Medicine, Nigde Omer Halisdemir University, Nigde, Turkey; 3Department of Biology, Faculty of Arts and Science, Duzce University, Duzce, Turkey; 4Department of Genetics, Aziz Sancar Experimental Medical Research Institute, University of Istanbul, Istanbul, Turkey; 5Department of Biology, Faculty of Arts and Science, Adiyaman University, Adiyaman, Turkey; 6Department of Biophysics, Faculty of Medicine, University of Istanbul, Istanbul, Turkey.

*These authors contributed equally to this work

Summary

Purpose: Adriamycin (ADR) is a commonly used anti-cancer drug. ADR has toxic effects on cardiomyocytes and leads to heart failure. However, the underlying mechanism(s) by which ADR causes heart failure is still not clarified exactly. The aim of present study is to investigate whether ADR-in-duced heart failure is mediated via HMGB1/TLR4 to initiate the apoptosis through MAPK/AMPK pathways.

Methods: H9c2 cell line was used to create four groups as a control, HMGB1 inhibition, ADR, ADR+HMGB1 inhibi-tion. Silencing HMGB1 was performed with specific small interfering RNA. ADR was used at 2 µM concentration for 36 and 48 hours. Protein and genes expressions, apoptosis was measured.

Results: Although ADR decreased AMPK, pAMPK, ERK1/2, pERK1/2, p38, JNK protein expression, ADR+HMGB1 inhi-bition led to change those protein expressions. The effect of silencing of HMGB1 prevented apoptosis induced by ADR in

the cells. HMGB1 caused changes a kind of posttranscrip-tional modification on the TLR4 receptor. This posttranscrip-tional modification of TLR4 receptor led to decreased AMPK protein level, but phosphorylated-AMPK. This alternation of AMPK protein caused enhancing of JNK protein, resulting from the decline of p38 and ERK protein levels. Eventually, JNK triggered apoptosis by a caspase-dependent pathway. The number of TUNEL positive and active caspase 8 cells at ADR was high, although HMGB1 silencing could decrease the cell numbers.

Conclusions: Inhibition of HMGB1 might prevent the lose of the cardiac cell by inhibition of apoptotic pathway, there-fore HMGB1 plays an essential role as amplifying on ADR toxicity on the heart by TLR4.

Key words: adriamycin, HMGB1, AMPK, TLR4, apoptosis, cardiac muscle cell

Introduction

Adriamycin (ADR), also called doxorubicin, is an anticancer drug widely used in the treatment of cancer since 1960s. However, it has adverse effects on many tissues such as liver, kidney and especially on the heart. Its toxic effect on the heart limits the use of this drug in the treatment of cancer patients.

ADR-induced cardiotoxicity and -heart failure can be observed shortly or years later after its adminis-tration [1]. Heart failure develops in 5-65% of can-cer patients treated with ADR [2,3]. Free reactive oxygen species (ROS) overproduction, disruption of cellular and mitochondrial Ca+2 homeostasis,

This work by JBUON is licensed under a Creative Commons Attribution 4.0 International License.

HMGB1 and adriamycin’s heart toxicity 555

JBUON 2020; 25(1): 555

induction of mitochondrial DNA breaks, impaired mitochondrial energy production, destruction of contraction proteins and cytoskeletal proteins in-cluding titin and dystrophin are suggested to be underlying mechanisms for ADR-induced cardio-toxicity. Each of these pathways can contribute to cardiac cell damage and ultimately apoptosis or ne-crosis [4,5]. However, the mechanisms that lead to the development of ADR-induced heart failure have not yet been fully elucidated. Therefore, an exact treatment strategy for the prevention or attenua-tion of the severity of ADR-induced heart failure in cancer patient has still not been developed. High-mobility group box 1 (HMGB1) is a well-preserved nuclear protein among species, acting as an alarm signal [6]. In addition to its nuclear role, HMGB1 also functions as an extracellular signal-ing molecule in several processes such as tissue regeneration, infection, cell differentiation, and tu-mor development [7]. HMGB1 is passively secreted from necrotic cells. Also, HMGB1 has been shown to be secreted under pathological conditions [8]. HMGB1 can bind to at least five different cell recep-tors such as advanced glycosylation end- product receptor (RAGE), TLR2, TLR4 [7]. It has been report-ed that HMGB1 is also released in oxidative stress and activates MAPKs (ERK, P38, JNK, SAPKs), NF-ΚB , calcium/calmodulin-dependent protein kinase (CaMKKβ) after interaction with its receptors [8, 9]. Moreover, HMGB1 exerts its cellular actions in infiltration related apoptosis mediated by lipopol-ysaccharide-induced myocardial insufficiency and ischemia-reperfusion injury via TNF-alpha [8]. A recent study indicates that HMGB1 also takes part in cardiotoxicity induced by ADR via autophagy [10]. The debate is going on the role of HMGB1 in apoptotic pathways in ADR-induced cardiotoxicity. A previous study indicated that cardiac apoptosis and dysfunction was diminished via blockage of HMGB1 in ADR exposed cardiac cells [8]. How-ever, another study showed that up-regulation of HMGB1 expression in ADR-induced cardiotoxicity resulted in increased expression of heat shock pro-tein beta 1 via attenuating cardiac apoptosis and mitochondrial dysfunctions [11]. In other words, ADR could trigger intracellular pathway(s) to am-ply its impact on tissues. According to scientific ranking, HMGB1 can be a good candidate for par-ticipation in ADR-induced toxicity. In the light of these knowledges, the underlying mechanisms medicated by HMBG1 in cardiac apoptosis after exposure of cardiac myocytes to ADR is not still clear. In this study, it is aimed to investigate the possible role of HMGB1 in ADR-induced apoptotic pathways in H9c2 cell line.

Methods

Cell culture and treatment

H9c2 cells was grown in Dulbecco’s Modified Ea-gle’s Medium (DMEM) containing 10% bovine serum albumin (BSA), 2 mM glutamine and antibiotics (100 IU/mL penicillin, 0.1 mg/mL streptomycin) in aseptic cell culture with various litter cell culture flasks at 37°C and 5% carbon dioxide. The cells in the flask started to work when they grew by 80%.

Experimental groups

In the present study, our hypothesis is that HMGB1 participate in ADR-induced cytotoxicity in cardiomyo-cytes. To test the hypothesis, HMGB1 gene was silenced by using siRNA [12] and four main experimental groups were created; (I) Control: DMEM, (II) siRNA: HMGB1 silenced (10 nM), (III) ADR: adriamycin treatment (2 μM), (IV) siRNA + ADR: adriamycin treatment and HMGB1 silenced. The siRNA inhibition of HMGB1 by lipofectamine and ADR treatment were done at the same time. If HMGB1 really involves ADR’s toxic effect on tissues, the synergetic effects could be evaluated when the silence of HMGB1 opted the beginning of ADR treat-ment. That is why, both treatments was preferred simul-taneously to start.

Transfection of H2c9 cells with HMGB1 siRNA

Cells were transfected with HMGB1 siRNA (Cat No: L-114889-00-0005; GE Healthcare Dharmacon) to sup-press HMGB1 gene expression in cells. Non-targeting siRNA (Cat No: D-001810-10-05; GE Healthcare Dharma-con) was used as a negative control. A hyperfect trans-fection agent (Cat No. 301702; Qiagen) was used for the transfection. Transfection optimization procedures were performed using different siRNA, transfection agent ra-tio. As a result of this trial, it was determined that 24 hours were not enough to silence the HMGB1 gene with siRNA. There was no significant difference in gene ex-pression levels between the control and treatment group (data not shown). Therefore, the relationship between HMGB1 and ADR were investigating depending on two periods: 36 and 48-hour interactions. After the time that is necessary for obtaining the efficient effect of siRNA was determined, the effective concentration was deter-mined. For this purpose, siRNA was administrated at 10, 25 and 50 nM concentrations by using 6-μL transfection agent. The ratio of HMGB1 gene suppression was 82% when 10 nM siRNA was used, 79% when 25 nM siRNA was used, and 85% when 50 nM siRNA was used. We obtained that there was no significant difference between the concentrations of 10 and 50 nM siRNA. As a result, it was decided to use 10 nM siRNA in concentration.

Determination of gene expression

RNA was isolated from cells and cDNA was synthe-sized and then gene expression levels were determined by Quantitative RT-PCR.

RNA isolation

The cultured cells were first centrifuged at 1500

HMGB1 and adriamycin’s heart toxicity556

JBUON 2020; 25(1): 556

rpm and the cells remaining at the bottom were washed 2 times with PBS after the medium was discarded. Af-ter the supernatant was poured, RNA was isolated from the remaining cells using the Roche RNA isolation kit according to the protocol. The RNA concentration and purity were calculated by taking 1 μL of the samples at the 260 nm wavelength by a Nanodrop device. It was also checked by running on a 1% gel.

RT-PCR

cDNA was synthesized using 1 μg of total RNA in 20 μL volume and random primers using the Roche Transcriptor cDNA synthesis kit according to the manu-facturer’s protocol.

Analysis of gene expression by real-time quantitative PCR

Real-time quantitative PCR (Light Cycler, Roche Di-agnostics GmbH, Germany) was used for quantitation. Primers for quantitative PCR were designed. For the PCR reaction, Light Cycler 480 SYBR Green Master Kit (Roche, Mannheim, Germany) and 250 ng cDNA sample were used. For the normalization of RNA quality and quantity in the samples, beta-actin was used as a control. PCR was run at 95ºC for 60 second (1 cycle) after pre-liminary denaturation, at 95ºC for 10 seconds, at 62ºC for 5 seconds, and at 72ºC for 20 seconds (45 cycles). After the amplification, the program applied for melting curve analysis was 0 seconds at 95ºC, 10 seconds at 65ºC and 0 seconds (1 cycle) at 95ºC. For each PCR reaction, water was used as a negative control instead of cDNA. Serial dilutions were prepared for each standard and each dilu-tion and each sample was run in duplicate.

Apoptotic cell count

TUNEL kit (Roche, Cat no: 11684795910) was used to determine the number of apoptotic cells. 150,000 cells were plated on glass coverslips for this purpose. Experi-mental groups were formed by applying ADR and/or siR-NA for 36- and 48-h to H9c2 cells, which were expected to adhere to glass for 24 h. DMEM is applied to control and negative control groups. Following the establish-ment of the experimental groups, the manufacturer’s protocol was followed. Briefly, freshly prepared 2% para-formaldehyde was added to 2 mL coverslips and incubat-ed for 40 min and fixation of the cells was achieved. Para-formaldehyde was removed by washing with phosphate buffer (PBS) 3 times for 2 min. Then, permeabilization (expansion of cell pores) with 0.02% Triton X-100 deter-gent was performed on ice for 10 min to allow easy entry of the dye in TUNEL kit into the cell. After, washing was performed 3 times with PBS for 2 min each. The labeling solution and the enzyme solution were mixed to freshly prepared master mixture. Enzyme-free labeling solution was used for negative control. Incubation was made in the dark and humidified atmosphere for 80 min. Finally, in the dark, the washing process was again carried out 3 times for 2 min each. Field scanning was performed using immunofluorescence microscopy at 20x, 40x and 100x objectives. The number of these red-stained nuclei was counted using the ImageJ program.

Determination of active caspase 8 by immunofluorescence method

In this analysis, IETD-FMK (FITC-IETD-FMK), a FITC-linked caspase 8 (Cas-8) inhibitor, was used as a marker. FITC-IETD-FMK is non-toxic and binds irrevers-ibly to activated Cas-8 found in apoptotic cells. Active Cas-8 was determined by immunofluorescence method according to the manufacturer’s protocol (Roche). Briefly, for each coverslip, 1 μL FITC-IETD-FKM was added to 500 μL of DMEM. Cells were incubated for 75 min in a 37°C incubator in FITC medium. Following the comple-tion of the incubation period, washing was carried out twice using a washing solution for 5 min each. Then, the samples were left to dry in the dark at room temperature for 10 min. Then, samples were scanned and imaged at 20x, 40x and 100x by using a fluorescence microscope.

Western blot analysis

The cell lysate proteins were loaded into the wells of the gels in such a way that 100 μg of protein was uniformly sampled from the samples to be separated into bands according to their molecular weights by elec-trophoresis. After loading, the proteins were run in gel electrophoresis (Invitrogen) for approximately one hour under 130 volts. Proteins carried in gel electrophoresis were transferred to the membrane using the iBlot device (Invitrogen). A signal enhancer kit (Super SignalTM, Inv-itrogen) was applied in order to make the protein bands to be obtained clearer and prominent before proceeding with the standard western blot protocol. The membrane was sacrificed with a solution containing BSA (5%) and blocked for 1 h at room temperature with an orbital mixer to minimize the nonspecific binding between membrane and antibody. Each of the primary antibod-ies (AMPK, TLR4, HMGB1, P38, SAPK / JNK, ERK 1/2, Cytochrome C, AIF, Cas-3, APAF-1, β-actin, PAMPK, PP38, PJNK, PERK 1/2) was diluted according to the manufac-turer’s protocol and the enhancer was added as blocking buffer. Membranes were then incubated with primary antibodies with orbital mixer at +4°C overnight. PVDF membranes were washed three times with PBS-T for 10 min. PVDF membranes were incubated in second-ary antibody (anti-rabbit and anti-mouse, 1/3000) and blocking buffer (5% BSA + PBS-T) for 1 h at room tem-perature using an orbital mixer. Then, membranes were washed three times for 10 min with PBS-T solution. Al-kaline phosphatase (BCIP / NBT) was used for imaging. Targeted proteins were identified using markers with standard molecular weights. Then, protein bands were photographed and analyzed with the Image J program.

Statistics

Data were evaluated using SPSS 21. The distribu-tion of the data was determined by Kolmogorov-Smirnov and Levene tests. In the evaluation of the significance of the difference between the averages of more than two groups; normal distribution conditions, one-way ANOVA followed by Tukey test as a post-hoc test was performed. The significance level was accepted as p<0.05. Values are given as mean ± standard error (mean ± SE).

HMGB1 and adriamycin’s heart toxicity 557

JBUON 2020; 25(1): 557

Results

Gene expression change in 36 hours

We explored the role of ADR in apoptosis via HMBG1 gene. As shown in Figure 1, the finding suggested that ADR treatment resulted in a three-fold reduction of HMGB1 gene expression. Next, to detect the role of HMGB1 in ADR-induced cardiac apoptosis, we silenced HMGB1 gene in control and ADR treated cardiomyocytes. Silencing of HMGB1 gene by siRNA resulted in a 3.5-fold reduction of HMGB1 gene expression compared to the control group. However, HMGB1 gene expression was sup-pressed seven-fold in cardiomyocytes treated with ADR and siRNA. ADR treatment led to a two-fold increase in TLR4 gene expression, 3.5-fold reduction in AMPK, a two-fold increase in ERK1 and 1.7-fold decrease in ERK2 gene expression and a five-fold in-crease in cytochrome-C gene expression. Silencing of HMGB1 gene resulted in a 0.5-fold increase in TLR4, 0.5-fold reduction in AMPK, 0.2-fold reduc-tion in ERK1 and 0.1-fold decrease in cytochrome-C gene expression. Treatment of cardiomyocytes with ADR and siRNA led to a three-fold increase in TLR4, a 4.5-fold decrease in AMPK, a 2.2-fold increase in ERK1 and 3.8-fold decrease in ERK2 and a seven-fold increase in cytochrome-C gene expression.

Gene expression change in 48 hours

As shown in Figure 1, siRNA transfection to cardiomyocytes resulted in a 2.5 fold reduction in HMGB1 gene expression. ADR treatment cause HMGB1 gene expression to decrease 1.1-fold. HMGB1 gene expression was reduced three-fold in cardiomyocytes treated with ADR and siRNA. ADR treatment resulted in a one-fold increase in TLR4, a 1.6-fold increase in ERK1 and 1.2-fold increase in ERK2. In addition, AMPK and cytochrome-C gene expression were almost suppressed in ADR treated cardiomyocytes. The silencing HMGB1 gene by

siRNA led to a 0.3-fold increase in TLR4, a 0.3-fold decrease in AMPK, a 0.5-fold decrease in ERK1 and 1.2-fold increase in ERK 2 and one-fold increase in cytochrome-C gene expression. Co-treatment of car-diomyocytes with ADR and siRNA resulted in a five-fold increase in TLR4, a 1.4-fold increase in ERK1 and 1.3-fold increase in ERK2 and 1.3-fold increase in cytochrome-C gene expression. In addition, AMPK gene expression was completely suppressed in car-diomyocytes treated with ADR and siRNA.

Protein expression HMGB1 protein

The silencing of HMGB1 gene resulted in a de-crease in expression of HMGB1 protein at 36h and 48h (Figure 2a). The protein expression of HMBG1 showed a significant enhance in ADR treated cardio-myocytes compared to control cell at 36h and 48h. Co-treatment of cardiomyocytes with ADR and siR-NA caused HMGB1 protein expression to decrease at both exposed time.

TLR4 receptor protein

TLR4 protein expression in the ADR group at 250 kDa decreased compared to the others, except siRNA+ADR group (Figure 2a). Although TLR4 pro-tein expression in 100 kDa in siRNA+ADR attenu-ated, there was seen a new band at around 150 kDa. Moreover, the new band at 150 kDa in ADR group was seen, but the expression of new 150 kDa in ADR was higher than the siRNA+ADR group in both ex-posure time.

AMPK-α1 protein

ADR treatment attenuated the protein expres-sion of AMPK-α1 when compared to control and siR-NA at 36h and 48h treatment (Figure 2a). AMPK pro-tein expression increased in cardiomyocytes treated with ADR and siRNA compared to ADR treatment at 36h and 48h treatment.

Figure 1. The mRNA expressions in the groups. Con: control group; siRNA: HMGB1 silence group; ADR: Adriamycin group; siRNA+ADR: HMGB1 silence plus adriamycin group. Data are expressed relative to that of β-actin.

HMGB1 and adriamycin’s heart toxicity558

JBUON 2020; 25(1): 558

Figure 2. The effect of HMGB1 on protein expression in H9c2 cells treated with adriamycin. a) The main hypothesis protein expression after 36h and 48h treatment. b) The mitogen-activated kinases and its phosphorylated protein expressions in groups after 36 h and 48 h treatment. c) Some apoptotic proteins expression in groups after 36 h and 48 h treatment. Con: control group; siRNA: HMGB1 silence group; ADR: Adriamycin group; siRNA+ADR: HMGB1 silence plus adriamycin group (n=3 independent experiments in groups).

HMGB1 and adriamycin’s heart toxicity 559

JBUON 2020; 25(1): 559

Phosphorylated AMPK-α1 protein

It was interesting that the silencing of HMGB1 gene by siRNA did not change the protein expres-sion of pAMPK compared to control (Figure 2a). However, ADR treatment increased pAMPK protein expression. The co-treatment of cardiomyocytes with ADR and siRNA decreased pAMPK protein ex-pression when compared to ADR.

Phosphorylated AMPK-α1

Interestingly, western blotting showed pAMPK-α1 bands at 65 and 50 kDa. siRNA increased the protein expression of pAMPK-α1 at both 65 and 50 kDa compared to control (Figure 2a). However, ADR decreased the protein expression of pAMPK-α1 at 65 kDa though increased it at 50 kDa in contrast to control. pAMPK-α1 protein expression was found to increase at 65 kDa and decrease at 50 kDa in car-diomyocytes treated with ADR and siRNA compared to ADR.

SAPK / JNK protein

ADR decreased SAPK/JNK protein expression at both exposed duration although there was no change in SAPK/JNK protein expression in control and siRNA groups (Figure 2b). SAPK/JNK protein ex-pression increased in cardiomyocytes treated with ADR and siRNA compared to ADR at both exposed duration.

Phosphorylated JNK protein

The silencing of HMGB1 gene by siRNA alle-viated phosphorylated JNK (pJNK) protein levels compared to control at 36h and 48h treatment (Fig-ure 2b). Although ADR treatment attenuated pJNK protein levels compared to the control group, the protein expression levels of pJNK were higher in cardiomyocytes treated with ADR and siRNA than that in the ADR group at 36h and 48h treatment.

P38 protein

The silencing of HMGB1 by siRNA elevated P38 protein expression compared to control at 36h and 48h (Figure 2b). Although ADR decreased P38 protein expression, the co-treatment with ADR and siRNA increased P38 protein expression compared to ADR alone at 36h and 48 h.

Phosphorylated P38

Phosphorylated P38 (pP38) protein expression was lower in siRNA group compared to the control group at 36h treatment (Figure 2b). ADR treatment decreased pP38 protein expression although ADR administration with siRNA increased pP38 protein expression compared to ADR at 36h and 48h.

ERK 1/2 protein

siRNA administration increased ERK 1/2 pro-tein expression compared to control. ADR decreased ERK1/2 protein expression compared to control. Co-treatment with ADR and siRNA increased ERK1/2 protein expression compared to ADR (Figure 2b).

Phosphorylated ERK 1/2

Phosphorylated ERK 1/2 protein (pERK 1/2) in siRNA groups was more apparent than that in the control group at 36 and 48h treatments. ADR treatment decreased pERK 1/2 protein expression in contrast to control groups at 36 h but increased at 48h treatment. The treatment of cardiomyocytes with ADR and siRNA resulted in a decrease in the expression of pERK1/2 protein when compared to ADR at 48h treatment (Figure 2b).

Cytochrome-C

Cytochrome-C is one of the most important regulatory proteins in the initiation of the apop-totic pathway. Therefore, we explored the change in protein expression of cytochrome-C. The result indicated that there were no significant changes in cytochrome-C protein expression among groups at 36h treatment. But, the protein expression of cyto-chrome-C at 48 h in the ADR group decreased com-pared to the others (Figure 2c).

Apoptosis-inducing factor

Apoptosis-inducing factor (AIF) is an important protein activating the apoptotic pathways, indepen-dently from caspases. AIF protein expression was found to be lower in ADR and co-treatment with ADR and siRNA compared to control at 36h treat-ment (Figure 2c). But there is no difference between groups at 48 h treatment.

Apoptotic peptidase activating factor 1

siRNA did not change the protein expression of apoptotic peptidase activating factor 1 (APAF-1) compared to control (Figure 2c). ADR increased the expression of APAF-1 protein at 36 and 48h treat-ment although the protein levels in co-treatment with ADR and siRNA was found to be lower in the ADR group at both duration.

Caspase-3

siRNA downregulated caspase-3 (Cas-3) protein in contrast to control at 36 and 48h treatments. ADR upregulated protein expression of Cas-3, although the protein expression level of Cas3 was lower in cardiomyocytes co-treated with ADR and siRNA than that in ADR at both duration (Figure 2c).

HMGB1 and adriamycin’s heart toxicity560

JBUON 2020; 25(1): 560

Figure 4. Apoptotic cell counts in H9c2 cells treated with adriamycin after 36 h and 48 h treatment. Con: control group; siRNA: HMGB1 silence group; ADR: Adriamycin group; siRNA+ADR: HMGB1 silence plus adriamycin group. Different let-ters indicate significant differences between groups (p<0.05).

Figure 3. Active Caspase-8 in in H9c2 cells treated with adriamycin after 36 h and 48 h treatment. First number of each alphabet represents 20x and its second number represents 100x magnification. A1-A2: Control group, B1-B2: siRNA group, C1-C2: Adriamycin group, D1-D2: HMGB1 silence plus adriamycin group. Although adriamycin enhanced active Caspase-8 number, HMGB1 silencing caused decrease of active caspase-8 number.

HMGB1 and adriamycin’s heart toxicity 561

JBUON 2020; 25(1): 561

Active Caspase 8 analysis after 36- and 48-hour treatments

After 36h, ADR led to an increase in the number of active Cas-8 (apoptotic cell) compared to control (Figure 3). The administration of siRNA to cell re-sulted in a decreased number of apoptotic cell in ADR-induced cell in contrast to the ADR group. Similarly, after 48h the cells exposed to ADR had intense active Cas-8. The silencing of HMGB1 gene by siRNA resulted in decreased activity of active cas-pase in ADR administered cells, supporting the role of HMGB1 gene in ADR-induced apoptotic pathways in cardiomyocytes.

TUNEL positive cell number after 36-hour and 48-hour treatments

After 36h and 48h, the number of TUNNEL positive cells significantly increased in ADR injected cells compared to control and siRNA groups. The

silencing of HMGB1 gene by siRNA resulted in de-creasing number of TUNNEL positive cells in con-trast to the ADR group (Figures 4 and 5).

Discussion

The most important findings of the present study are (I) ADR exerts its undesired effect on cardiac cells via HMGB1; (II) HMGB1 causes a posttranslational change in TLR4 receptor; (III) TLR4 activates MAPK, resulting in apoptosis by the caspase-dependent pathway. ADR causes a cytotoxic effect on noncancer-ous tissue, especially on heart. Therefore, it is im-portant to clarify the underlying mechanism of its toxic effect. The idea of the present study was that ADR may possibly participate in its toxic effect on noncancerous tissue through intracellular path-ways. One of candidate agents might be HMGB1

Figure 5. Apoptosis in H9c2 cells treated with adriamycin after 36h and 48h treatment. First number of each alphabet represents 20x and its second number represents 100x magnification. A1-A2: Control group, B1-B2: siRNA group, C1-C2: Adriamycin group, D1-D2: HMGB1 silence plus adriamycin group. Although adriamycin enhanced Tunel-Positive cell num-bers, HMGB1 silencing caused decrease Tunel-positive cell numbers.

HMGB1 and adriamycin’s heart toxicity562

JBUON 2020; 25(1): 562

because it is one of the damage-associated molecu-lar pattern (DAMP) [13]. In the other words, HMGB1 acts as a molecule associated with extracellular damage [14]. It acts as an extracellular signaling molecule, especially in acute and chronic inflam-mation. While the excess of HMGB1 plays a patho-logical role in cardiovascular diseases, it may be beneficial to decrease the level of HMGB1. So, ADR has been reported to increase the protein levels of HMGB1 [10,15]. The mRNA data of HMGB1 in the present study represented to decrease, whereas the significant elevation on HMGB1 protein levels in ADR. The silence of HMGB1 caused to decrease the protein levels. Thus, HMGB1 gene silencing or its effects may not only be effective in non-cancerous tissues such as heart muscle but may also be ef-fective in the treatment of cancer in the preven-tion or reduction of the resistance to chemothera-peutic drugs such as ADR. Nevertheless, it should not be forgotten that HMGB1 is indispensable for life because it is indicated that the life span of HMGB1-/- silenced mice is shortened [16]. HMGB1 can bind cell receptors such as RAGE, TLR2, TLR4. In our study, we suggested that HMGB1 could participate in ADR’s toxic effect on cardiomyocytes via TLR4. In this context, the re-sults obtained from the studies made especially for our hypothesis are exciting. According to the result, ADR cause to increase HMGB1 levels, more-over TLR4 probably involve in HMGB1 effects on cardiomyocytes because TLR4 protein expression has some alternation. The post-translation modifi-cation of TLR, such as glycosylation, is so crucial to participate in the receptors localization, signal-ing capacities, and assembly [17]. Glycosylation is reported to one of most seen and complexes post-translation modification complying to covalent at-tachment of oligosaccharide chain to -NH2 group of asparagines (N-glycosylation). It is believed to this glycoprotein generally located in cell mem-brane or organel lumens, thus participated in many physiologic processes, including immunologic re-action [18]. Moreover, TLR4 is one of α-2-6- and α-2,3-sialylated glycoproteins like MD2. So, when sialyl residues in TLR4 is removed, cells could in-crease NF-κB activation, cytokine releasing and TLR4 dimerization [17]. Also, sialidase Neu 1 is related to TLR2, TLR3 besides TLR4, enhancing following agonist trigger [17]. TLR4 is reported to possess N-glycosylated proteins, meaning that core fucose, α1,6-linked fucosylation of the innermost N-acetylglucosamine residue of N-linked glycan, can affect TLR4-dependent signaling [19]. Also, fatty acid modification of TLR4 was reported to be important for its function [20]. The higher bands of the protein in western blot are usually reported to

post-translational modification [21-23]. According to the antibody manufacturer info used in the pre-sent study, the higher band of TLR4 than 170 kDa means to the heavily glycosylated or complexed form or TLR4 despite the heat of cell lysate for 10 minutes at 100°C in the present study. Unfortu-nately, we could not able to confirm the posttrans-lational modification of TLR4 which is the limita-tion of the current study, excepting our Western blot result. The main mechanism underlying heart failure is cardiac cell loss due to apoptosis, resulting in car-diomyocytes death and severe dysfunction in cardi-ac contractility [8,11]. However, the mechanism of this apoptotic process is still a mystery. So, we in-vestigated intracellular pathways of ADR/HMGB1/TLR4 pathways through MAPK/AMPK. AMPK is a regulator and chief actor in cell energy and nutri-tion signaling. This kinase is generally activated in cell energy stress induced by either physiological condition such as exercise that increases the in-tracellular calcium, reactive oxygen, and nitrogen levels or pathological conditions such as ischemia that elevate the ratio of AMP/ATP, ADP/ATP [24]. AMPK is also associated with MAPK [24]. MAPK is a tightly conserved serine/threonine protein ki-nase family that transitions between species and plays a role in many processes such as infiltration, immune, differentiation, cell death, and survival. This family has three members, ERK, JNK and P38 [16]. ERK1/2 activated by all kind of stress in a cell, plays a role in the survival of the cell (anti-apop-totic). JNK has a pro-apoptotic effect. In our study, we propose as a hypothesis that ADR alleviated the protein expression of pro-apoptotic JNK protein via decreasing ERK1/2 protein level. In addition, it was indicated that ADR suppresses AMPK in the heart, but increases MAPK signaling pathway. Because ADR suppresses AMPK, ATP production capacity is decreasing. While ADR activates AMPK directly by impairing its energy balance, it also suppresses AMPK indirectly by its oxidative stress and its ge-notoxic effect [24]. A recent study suggests that ADR activates AMPK and, conversely, reduces P38 and mTOR [25]. However, AMPK is reported to activate via P38 and ASK1. Both P38 and ASK1 are activated via extensive cellular responses to stress. Inhibition of ASK1 also phosphorylates P38 and reduces oxidative damage. AMPK inhibits P38 and ASK via inactivating them. This information demonstrates that ADR can affect AMPK through many intracellular signaling pathways. Therefore, this may be a reason for the difference in the result of AMPK expression. It is a sign of the strength to resolve the mechanism of changes in AMPK. In our study, ADR reduced the protein expression

HMGB1 and adriamycin’s heart toxicity 563

JBUON 2020; 25(1): 563

levels of P38 and PP38 protein. In addition, the silencing of the HMGB1 gene caused an increase in the level of AMPK protein. A study suggests that AMPK may be protective against cell damage from ADR-induced renal podocytes. It has also been noted that oxidative stress activates AMPK [26]. There are many sources in the literature that oxi-dative stress is an important mechanism in heart failure caused by ADR [27-33]. Since AMPK is an important protein for cell function, it interacts with many intracellular pathways and mediators. One of these pathways was recently defined. In a previ-ous study, ADR has been reported to inhibit AMPK with time. Interestingly, ADR has also been shown to increase ERK1/2 phosphorylation in this cancer line. In addition, ADR has induced disruption of Caspase-3 (Cas-3). It is emphasized that this is an important mechanism in the development of resist-ance to ADR [34]. The heart often consumes ATP, and most are used in muscle contraction. Adverse effects of ADR on heart energy metabolism are me-diated by impairing oxidative phosphorylation, mi-tochondrial electron transport system and AMPK functions [35]. AMPK is one of the most important proteins in the maintenance of cellular energy. Thus, AMPK plays a major role in survival [36] and/or apoptotic pathways [37]. Because AMPK is an ac-tive cascade that requires energy in apoptosis, we explored the change in protein expression level of AMPK-α1. Interesting knowledge on AMPK is in-dicated that TAK1 is one of the upstream activators of AMPK. JNK and S6 kinase leads to an increase in NF-κB expression via IKK. TAK1-P38 in MH60 cell line was reported to trigger apoptosis by BMP-2. Moreover, blocking of TAK1 is reported to enhance susceptibility to ADR in HeLa (adenocarcinoma) and mouse embryonic fibroblasts (MEFs) cell lines [38]. It was also reported that HMGB1’s inhibition led to inhibit TLR4/MyD88/ NF-κB pathways in brain ischemia/reperfusion [39]. Caspases are intracellular cysteine-aspartic protease groups, which mediate apoptosis. Cas-pases are found as inactive proenzymes and are activated by proteolytic enzymes. Cas-8 is thought to be an apoptotic and cytokine-mediated cell death receptor [16]. In our study, the amount of active Cas-8 was determined in live cardiomyocytes be-cause of the ability of HMGB1 to act as a cytokine. Suppression of the HMGB1 gene resulted in a de-crease in the amount of ADR-induced active Cas-8. Cas-9 is effective in both the mitochondrial death pathways and the passage of cytochrome-c from mitochondria to cytoplasm. Cas-9 is also effective in the mitochondrial death pathway and is effective in the passage of cytochrome-c from mitochondria to cytoplasm. Cas-3 and Cas-7 serve as an enhanc-

er in this apoptotic process initiated by Cas-8 and Cas-2 [16]. In the present study, we also examined the protein expression level of Cas-3. The results showed that blocking HMGB1 inhibits apoptosis via attenuating Cas-3 and Cas-8 levels, which is an indication that apoptosis occurs in the caspase-de-pendent pathway. However, although caspase activ-ity plays a critical role in apoptotic cell death path-way, it has been described that apoptosis-inducing factor (AIF) and endonuclease G plays an important role in the translocation from mitochondria to the nucleus, independently from the caspase, which results in DNA breaks [16]. In the present study, it was proposed that Cas-3 is important in ADR-induced heart failure and apoptotic cell death oc-curs via caspase-mediated pathways depending on the observation that there was no change in the expression of AIF protein. HMGB1 not only regu-lates endonuclease activity, but it also becomes a part of apoptotic bodies [16]. With the light of this information, we built the main hypothesis of our work on HMGB1. HMGB1 take part in cell migration via RAGE and mediates inflammation and gene transcription through TLR4 and P53. The pro-inflammatory ef-fects of HMGB1 are mediated by P38, ERK, JNK. The accumulation of HMGB1 in the extracellular fluid causes toxic effects and tissue damage such as apoptosis, necrosis, and cell death. The extracel-lular HMGB1 causes the formation of giant mito-chondrial vacuoles and rapidly mitochondrial DNA break by entering mitochondria in an endocytosis-independent manner. HMGB1’s entry into the mi-tochondria is not via TLR-2, TLR4 or not RAGE signaling pathways. However, it is thought that the activation of JNK through ROS is effective in this process [16], which support ADR-induced apoptotic pathways in the present study. This information enhances the importance of the result of the pre-sent study when one of the intracellular targets of ADR is combined with knowledge of mitochondria [27-32]. It is stated that the amount of HMGB1 in-creases in cancer and decreases in aging [16]. JNK is a kinase involved in cell apoptosis and inflammation. Cardiac cells increase HMGB1 pro-duction and release in stress conditions such as sepsis and ischemia/reperfusion. After being re-leased from the cell, HMGB1 binds to its receptor, leading to pro-inflammatory effects and resulting in apoptotic cell death and heart failure. One of the previous study suggested that ADR increased the production of HMGB1 in the heart and triggered apoptosis [8]. JNK, a proapoptotic kinase, is thought to play an important role in apoptosis in the heart. The main mechanism of the JNK activation is oxi-dative stress. It also indicates that ADR increases

HMGB1 and adriamycin’s heart toxicity564

JBUON 2020; 25(1): 564

the expression of HMGB1 via JNK in the mentioned study. Therefore, it was found that the amount of HMGB1 decreased when the effect of JNK was sup-pressed, which demonstrate JNK’s importance in the apoptotic pathway caused by ADR. It is em-phasized that either genetic (in JNK or TLR4 gene silencing) or pharmacologically suppression of the action of HMGB1 can not completely prevent the toxic effect of ADR seen in the heart [8]. Our result supports the mentioned study. The silencing of the HMGB1 gene significantly reduced apoptotic cell death. We presumably believe that some of these toxic effects of ADR may be mediated through the interaction of HMGB1 specifically with the TLR4 receptor.

Conclusion

The undesired effect of ADR on heart tissue, noncancerous tissue, participate in HMGB1/TLR4 axis through MAPK/AMPK pathways, resulting in triggering in apoptotic cell death. The most inter-

esting results in the current study was seen on some posttranslational modification of TLR4. It is limitation in the study not to deeply analyze the modification. It is suggested to evaluate the modi-fication of TLR4 by future studies.

Acknowledgements

We really appreciate Dr. Belma Turan (Depart-ment of Biophysics, University of Ankara School of Medicine) for a kind gift of H9c2 cells (embryonic rat heart cell line).

Funding

This study was supported by The Scientific and Technological Research Council of Turkey (TUBI-TAK) as a Project No: 114S118.

Conflict of interests

The authors declare no conflict of interests.

References

1. Bristow MR, Thompson PD, Martin RP, Mason JW, Bill-ingham ME, Harrison DC. Early anthracycline cardio-toxicity. Am J Med 1978;65:823-32.

2. Minami M, Matsumoto S, Horiuchi H. Cardiovas-cular side-effects of modern cancer therapy. Circ J 2010;74:1779-86.

3. Horacek JM, Vasatova M, Pudil R et al. Biomarkers for the early detection of anthracycline-induced cardiotox-icity: current status. Biomed Pap Med Fac Univ Palacky Olomouc Czech Repub 2014;158:511-7.

4. Simunek T, Sterba M, Popelova O, Adamcova M, Hrdina R, Gersl V. Anthracycline-induced cardiotoxicity: over-view of studies examining the roles of oxidative stress and free cellular iron. Pharmacol Rep 2009;61:154-71.

5. Minotti G, Menna P, Salvatorelli E, Cairo G, Gianni L. Anthracyclines: molecular advances and pharmacolog-ic developments in antitumor activity and cardiotoxic-ity. Pharmacol Rev 2004;56:185-229.

6. Tsoyi K, Jang HJ, Nizamutdinova IT et al. Metformin in-hibits HMGB1 release in LPS-treated RAW 264.7 cells and increases survival rate of endotoxaemic mice. Br J Pharmacol 2011;162:1498-1508.

7. Wu L, Yang L. The function and mechanism of HMGB1 in lung cancer and its potential therapeutic implica-tions. Oncol Lett 2018; 15:6799-6805.

8. Yao Y, Xu X, Zhang G, Zhang Y, Qian W, Rui T. Role of HMGB1 in doxorubicin-induced myocardial apo-ptosis and its regulation pathway. Basic Res Cardiol 2012;107:267.

9. Tang D, Kang R, Zeh HJ, 3rd, Lotze MT. High-mobility group box 1, oxidative stress, and disease. Antioxid Re-dox Signal 2011;14:1315-35.

10. Luo P, Zhu Y, Chen M et al. HMGB1 contributes to adriamycin-induced cardiotoxicity via up-regulating autophagy. Toxicol Lett 2018;292:115-22.

11. Narumi T, Shishido T, Otaki Y et al. High-mobility group box 1-mediated heat shock protein beta 1 ex-pression attenuates mitochondrial dysfunction and apoptosis. J Mol Cell Cardiol 2015;82:1-12.

12. Qin WD, Mi SH, Li C et al. Low shear stress induced HMGB1 translocation and release via PECAM-1/PARP-1 pathway to induce inflammation response. PLoS One 2015;10:e0120586.

13. Pellegrini L, Foglio E, Pontemezzo E, Germani A, Russo MA, Limana F. HMGB1 and repair: focus on the heart. Pharmacol Ther 2019;196:160-82.

14. Chen Q, Guan X, Zuo X, Wang J, Yin W. The role of high mobility group box 1 (HMGB1) in the pathogenesis of kidney diseases. Acta Pharm Sin B 2016;6:183-8.

15. Zhang H, Lu X, Liu Z, Du K. Rosuvastatin reduces the pro-inflammatory effects of adriamycin on the ex-pression of HMGB1 and RAGE in rats. Int J Mol Med 2018;42:3415-23.

16. Kang R, Chen R, Zhang Q et al. HMGB1 in health and disease. Mol Aspects Med 2014;40:1-116.

17. Leifer CA, Medvedev AE. Molecular mechanisms of regulation of Toll-like receptor signaling. J Leukoc Biol 2016;100:927-41.

HMGB1 and adriamycin’s heart toxicity 565

JBUON 2020; 25(1): 565

18. Sachse R, Wüstenhagen D, Šamalíková M, Gerrits M, Bier FF, Kubick S. Synthesis of membrane proteins in eukaryotic cell-free systems. Engineer Life Sci 2013;13:39-48.

19. Iijima J, Kobayashi S, Kitazume S et al. Core fucose is critical for CD14-dependent Toll-like receptor 4 signal-ing. Glycobiology 2017;27:1006-15.

20. Rowe DC, McGettrick AF, Latz E et al. The myristoyla-tion of TRIF-related adaptor molecule is essential for Toll-like receptor 4 signal transduction. Proc Natl Acad Sci U S A 2006;103:6299-6304.

21. Choi BH, Chen C, Philips M, Dai W. RAS GTPases are modified by SUMOylation. Oncotarget 2018;9:4440- 50.

22. Oh Y, Chung KC. Small ubiquitin-like modifier (SUMO) modification of zinc finger protein 131 potentiates its negative effect on estrogen signaling. J Biol Chem 2012;287:17517-29.

23. Spence J, Gali RR, Dittmar G, Sherman F, Karin M, Fin-ley D. Cell cycle-regulated modification of the ribosome by a variant multiubiquitin chain. Cell 2000;102:67-76.

24. Gratia S, Kay L, Potenza L et al. Inhibition of AMPK signalling by doxorubicin: at the crossroads of the car-diac responses to energetic, oxidative, and genotoxic stress. Cardiovasc Res 2012;95:290-99.

25. Wang X, Wang XL, Chen HL et al. Ghrelin inhibits doxorubicin cardiotoxicity by inhibiting excessive autophagy through AMPK and p38-MAPK. Biochem Pharmacol 2014;88:334-50.

26. Gao K, Chi Y, Sun W, Takeda M, Yao J. 5’-AMP-activated protein kinase attenuates adriamycin-induced oxida-tive podocyte injury through thioredoxin-mediated suppression of the apoptosis signal-regulating kinase 1-P38 signaling pathway. Mol Pharmacol 2014;85:460-71.

27. Dursun N, Taskin E, Yerer Aycan MB, Sahin L. Sele-nium-mediated cardioprotection against adriamycin-induced mitochondrial damage. Drug Chem Toxicol 2011;34:199-207.

28. Guven C, Taskin E, Akcakaya H. Melatonin Prevents Mitochondrial Damage Induced by Doxorubicin in Mouse Fibroblasts Through Ampk-Ppar Gamma-De-pendent Mechanisms. Med Sci Monit 2016;22:438-46.

29. Taskin E, Dursun N. The protection of selenium on adriamycin-induced mitochondrial damage in rat. Biol Trace Elem Res 2012;147:165-71.

30. Taskin E, Dursun N. Recovery of adriamycin induced mitochondrial dysfunction in liver by selenium. Cyto-technology 2015;67:977-86.

31. Taskin E, Kindap EK, Ozdogan K, Aycan MB, Dursun N. Acute adriamycin-induced cardiotoxicity is exacerbat-ed by angiotension II. Cytotechnology 2016;68:33-43.

32. Taskin E, Ozdogan K, Kunduz Kindap E, Dursun N. The restoration of kidney mitochondria function by inhibition of angiotensin-II production in rats with acute adriamycin-induced nephrotoxicity. Ren Fail 2014;36:606-12.

33. Yapislar H, Taskin E, Ozdas S, Akin D, Sonmez E. Coun-teraction of Apoptotic and Inflammatory Effects of Adriamycin in the Liver Cell Culture by Clinopitolite. Biol Trace Elem Res 2016;170:373-81.

34. Zhang Y, Duan G, Feng S. MicroRNA-301a modulates doxorubicin resistance in osteosarcoma cells by tar-geting AMP-activated protein kinase alpha 1. Biochem Biophys Res Commun 2015;459:367-73.

35. Sturgeon K, Muthukumaran G, Ding D, Bajulaiye A, Ferrari V, Libonati JR. Moderate-intensity treadmill ex-ercise training decreases murine cardiomyocyte cross-sectional area. Physiol Rep 2015;3:E12406.

36. Bairwa SC, Parajuli N, Dyck JR. The role of AMPK in cardiomyocyte health and survival. Biochim Biophys Acta 2016;1862:2199-2210.

37. Hsu YC, Meng X, Ou L, Ip MM. Activation of the AMP-activated protein kinase-p38 MAP kinase pathway me-diates apoptosis induced by conjugated linoleic acid in p53-mutant mouse mammary tumor cells. Cell Signal 2010;22:590-9.

38. Aashaq S, Batool A, Andrabi KI. TAK1 mediates con-vergence of cellular signals for death and survival. Apoptosis 2019;24:3-20.

39. Zhang J, Xiao F, Zhang L et al. Alpha-Lipoic Acid Pre-conditioning and Ischaemic Postconditioning Syner-gistically Protect Rats from Cerebral Injury Induced by Ischemia and Reperfusion Partly via Inhibition TLR4/MyD88/ NF-kappaB Signaling Pathway. Cell Physiol Biochem 2018;51:1448-60.